Public health authorities encourage persons at risk for AIDS to seek voluntary HIV antibody testing in order to become motivated to make behavior changes that lessen the risk of transmitting the virus to others. The importance of counseling individuals after HIV testing is repeatedly stressed by public health authorities, although the standard counseling which most HIV-tested persons receive consists of only brief information about test meaning and steps needed to protect others. Studies have shown that HIV testing with brief informational counseling produces only partial reductions in subsequent rates of high-risk conduct and fails to alleviate symptoms of emotional distress. Further, literature in other health-behavior areas indicates that brief informational counseling alone is generally insufficient to promote meaningful, permanent change in longstanding risk behavior patterns. The purpose of this study is to evaluate alternative models for counseling HIV seropositive persons. The proposed study will compare three approaches for counseling persons who learn they are HIV-positive on measures of subsequent high-risk behavior and emotional/behavioral adjustment. One hundred-fifty homosexual and bisexual males with a history of high-risk behavior and who are HIV-positive after voluntary testing will be randomly-assigned to one of three intervention conditions: standard informational counseling alone as it is usually provided by alternative test sites (IC), IC plus participation in a support group, or IC plus a behavioral group intervention that trains self-management, assertiveness, social support, and anxiety management skills tailored to implementing changes which protect others from infection and promote effective stress coping. Before and after the intervention, and at 6-month followup intervals over 18 months, all subjects will be administered objective measures assessing the frequency of specific high-risk sexual activities and substance use patterns; will self-monitor similar behaviors for six-week periods; and will complete standardized inventories of depression, anxiety, anger, and AIDS risk knowledge. Data from subjects in the three intervention conditions will be compared using multivariate analyses to determine which type of intervention produces the greatest degree of short-term and long-term change in risk behavior and emotional distress. The results of the study will provide empirical data essential to the development of improved behavior change intervention strategies for counseling persons who are found to be HIV seropositive.